Actions
Bug #198
closedvalgrind report a context error on SSLParserTest01
Affected Versions:
Effort:
Difficulty:
Label:
Description
valgrind sometimes reports the following error with -O2 on a 64bit vmware box:
pablo@kdevel:/home/pablo/oisf/src# file .libs/suricata
.libs/suricata: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, not stripped
11613 at 0x5D6044: SSLParseClientRecord (app-layer-ssl.c:91)
11613 by 0x5B63EC: AppLayerDoParse (app-layer-parser.c:675)
11613 by 0x5B6F4D: AppLayerParse (app-layer-parser.c:879)
11613 by 0x5D7A10: SSLParserTest01 (app-layer-ssl.c:275)
11613 by 0x53A90A: UtRunTests (util-unittest.c:199)
11613 by 0x405E95: main (suricata.c:893)
11613
11613 ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 27 from 4)
Files
Updated by Victor Julien over 14 years ago
- Due date set to 06/30/2010
- Assignee changed from OISF Dev to Pablo Rincon
- Target version set to 1.0.0
- Estimated time set to 2.00 h
Updated by Pablo Rincon over 14 years ago
- File 0002-This-patch-for-app-layer-ssl-fix-the-bug-198-SSLPa.patch 0002-This-patch-for-app-layer-ssl-fix-the-bug-198-SSLPa.patch added
- Status changed from New to Resolved
It's really weird, but it seems that the way of initializing the input was not handled properly by gcc on -O2 and -O3. I changed it to avoid the usage of brackets as array, and directly use a string with hexa data, fixing the length, since it has \0's, and now it doesn't report ctx errors. The problem was at the last 2 bytes, after two 0x00.
Updated by Victor Julien over 14 years ago
- Status changed from Resolved to Closed
- % Done changed from 0 to 100
Applied, thanks Pablo.
Actions